Generative AI & LLMs
Context Engineering & Long Context
9 practice questions. Free questions open a full answer guide; the rest unlock with Pro.
- You have a long-running agent that accumulates conversation and tool output until it overruns its context window mid-task. Walk me through the techniques you'd use to keep it coherent over a long task without just truncating history.
- You append tool outputs and conversation history to the prompt on every step of an agent, and your token bill and latency are climbing. How would you bring both down without losing the context the agent actually needs?Go Pro
- What does it mean to treat the context window as a budget, and how would you decide what to include for a given request?Go Pro
- Why can't an LLM just remember everything if you put a huge amount of text in its context window? What goes wrong as the input gets very long?Go Pro
- Your model has a million-token context window, but accuracy drops when the key fact sits in the middle of a long input. Why does this happen, and how would you design around it?Go Pro
- For a chatbot that has a long conversation with a user, how would you keep the relevant context available without the prompt growing forever?Go Pro
- Your long-context LLM service is slow and expensive because every request resends a large, mostly-static instruction-and-context block. Without shrinking what the model sees, how would you cut latency and cost — and what changes about how you order and structure the prompt?Go Pro
- Models now advertise huge context windows, but teams report that quality drops as you fill them. Explain what's happening and how you'd manage what goes into the window for a long-running task.Go Pro
- Your agent needs to remember facts and decisions across many turns and sessions, but you can't keep the entire history in the context window. How would you design memory for it?Go Pro
Want questions matched to your role? Paste a job title, job description, or CV for a personalized set, or go Pro to unlock the full bank.